DOI: 10.1111/1471-0528.13057
General obstetrics
www.bjog.org
Cardiovascular disease risk is only elevated in
hypertensive, formerly preeclamptic women
NM Breetveld,a C Ghossein-Doha,a SMJ van Kuijk,b AP van Dijk,c MJ van der Vlugt,c
WM Heidema,d RR Scholten,d MEA Spaandermana
a
Department of Obstetrics and Gynaecology, Research School GROW, Maastricht University Medical Centre (MUMC), Maastricht, the
Netherlands b Department of Epidemiology, Maastricht University, Maastricht, the Netherlands c Department of Cardiology, Radboud
University Medical Centre (Radboudumc), Radboud, the Netherlands d Department of Obstetrics and Gynecology, Radboud University
Medical Centre (Radboudumc), Radboud, the Netherlands
Correspondence: Dr Ms NM Breetveld, Department of Obstetrics and Gynaecology, P. Debyelaan 25, 6229 HX Maastricht, PO Box 5800,
6202 AZ Maastricht, the Netherlands. Email n.breetveld@student.maastrichtuniversity.nl
Accepted 13 July 2014. Published Online 20 August 2014.
Objective To analyse the predicted 10- and 30-year risk scores for
cardiovascular disease (CVD) in patients who experienced
preeclampsia (PE) 5–10 years previously compared with healthy
parous controls.
Design Observational study.
Setting Tertiary referral hospital in the Netherlands.
Population One hundred and fifteen patients with a history of PE
and 50 controls. PE patients were categorised into two groups,
hypertensive (n = 21) and normotensive (n = 94), based on use of
antihypertensive medication, and next categorised into subgroups
based on the onset of PE: early-onset PE (n = 39) and late-onset
PE (n = 76).
Methods All participants underwent cardiovascular risk screening
5–10 years after index pregnancy. We measured body mass, height
and blood pressure. Blood was analysed for fasting glucose,
insulin and lipid levels. All participants completed a validated
questionnaire. The 10- and 30-year Framingham risk scores were
calculated and compared.
Main outcome measures Estimated Framingham 10- and 30-year
Results The overall 10- and 30-year CVD median risks weighing
subjects’ lipids were comparable between formerly PE women and
controls; 1.6 versus 1.5% (P = 0.22) and 9.0 versus 9.0%
(P = 0.49), respectively. However, hypertensive formerly PE
women have twice the CVD risk as normotensive formerly PE
women: 10- and 30-year CVD median risks were 3.1 versus 1.5%
(P < 0.01) and 19.0% versus 8.0% (P < 0.01), respectively. Risk
estimates based on BMI rather than lipid profile show comparable
results. Early-onset PE clustered more often in the hypertensive
formerly PE group and showed significantly higher 10- and 30year CVD risk estimates based on lipids compared with the lateonset PE group: 1.7 versus 1.3% (P < 0.05) and 10.0 versus 7.0%
(P < 0.05), respectively.
Conclusions Women who are hypertensive after preeclampsia,
have a twofold risk of developing CVD in the next 10–30 years.
Formerly PE women who are normotensive in the first 10 years
after their preeclamptic pregnancy have a comparable future
cardiovascular risk to healthy controls.
Keywords Cardiovascular risk, Framingham risk score,
hypertension, metabolic syndrome, preeclampsia.
risk scores for CVD.
Please cite this paper as: Breetveld NM, Ghossein-Doha C, van Kuijk SMJ, van Dijk AP, van der Vlugt MJ, Heidema WM, Scholten RR, Spaanderman MEA.
Cardiovascular disease risk is only elevated in hypertensive, formerly preeclamptic women. BJOG 2015;122:1092–1100.
Introduction
Preeclampsia (PE), a vascular pregnancy-related disorder
complicating 5–8% of all pregnancies1 is not only a major
cause of fetal and maternal morbidity and mortality2,3 but
also increases the risk for premature cardiovascular disease
(CVD) later in life.4,5 PE is diagnosed as new-onset hypertension after 20 weeks gestational age with proteinuria
(0.3 g/day).6 Depending on gestational age at delivery,
patients with a history of PE have about a two- to
1092
sevenfold risk of developing ischaemic cardiac disease
compared with healthy controls and an approximately
four-fold risk of developing chronic hypertension within
15 years after pregnancy.4 Cardiovascular disease is the
number one cause of death in women.7 It is not known
whether PE itself, as a gender-specific disorder, increases
this risk independently or through risk factors known to be
associated with both PE and CVD. Patients with a history
of PE exhibit more often constituents of the metabolic syndrome, e.g. insulin resistance, dyslipidaemia, hypertension,
ª 2014 Royal College of Obstetricians and Gynaecologists
Framingham risk score in formerly preeclamptic women
micro-albuminuria and obesity compared with patients
with a history of uncomplicated pregnancies.8–11 When
detected in time, these factors may be modifiable,12
emphasising the importance of cardiovascular follow-up in
these patients. Quantifying the person-specific CVD risk
could be useful in counselling these patients.
At present, there is no tailored risk score available to
predict cardiovascular disease in this specific population
(women with a history of PE). Several risk scores have been
proven to be of predictive value in other different risk populations. The Framingham risk score calculator is the most
widely used and a well validated strategy to estimate personalised risk.13–15 Although a few studies have analysed
the Framingham risk score in a population of former PE
women, these studies did not differentiate between the risk
calculator modelled on lipids or on BMI.16–18 Moreover,
these studies did not differentiate between the subgroups of
former PE patients, either by onset of disease or later postgestation chronic hypertension.16–18 The aim of this study
was to compare the predicted risk of cardiovascular disease
in the next 10 and 30 years as computed with the Framingham risk score calculator, between patients with a history
of PE and women who had a normotensive pregnancy in
the past (5–10 years after index pregnancy) and subsequently compare subgroups within the former PE group
based on onset of disease and/or whether hypertension has
developed. We hypothesise that (1) patients with a history
of PE have higher predicted Framingham CVD risk scores
compared with patients with a history of only uncomplicated pregnancies and that (2) the risk estimates vary
among the different subgroups of former PE patients.
Methods
The study protocol of this observational study was
approved by the Medical Ethics Committee of the Radboud
University Medical Centre (CMO: 2010/245). For this
study, women were included between 2010 and 2012.
Study population
Formerly preeclamptic women were recruited from a database of women who had preeclampsia and volunteered to
participate in a cardiovascular follow-up study program. PE
in index-pregnancy was diagnosed according to criteria
set the International Society of Hypertension in Pregnancy: new-onset hypertension, systolic blood pressure
(SBP) ≥ 140 mmHg and/or diastolic blood pressure
(DBP) ≥ 90 mmHg, after 20 weeks’ gestation and proteinuria exceeding 0.3 g/day.6,19 Early-onset PE was defined as
PE developing before 34 weeks’ gestation. [Correction added
on 5 December 2014, after first online publication: the
number of weeks of gestation has been changed from ‘35
weeks’ to ‘34 weeks’ in the preceding sentence.] Women
ª 2014 Royal College of Obstetricians and Gynaecologists
who had completed the 5- to 10-year postpartum cardiovascular risk screening were eligible for analysis in the current
study. Formerly preeclamptic women who participated in a
previous study were invited to participate in this study by
mail. Initially, former PE women were recruited at the 6week postpartum screening by a clinician, and were seen at
1 year postpartum, and invited to participate in this study
approximately 5 years later. The catchment area of this
study population was in the east side of the Netherlands in
which the socioeconomic status is reported to be average
with a low prevalence of immigrants. Controls were
recruited by advertisement in local newspapers, schools and
childcare centres in the same area. Women in the control
group had to be between 25 and 45 years old, and to have
had their first pregnancy 5–10 years earlier.
Pregnancy charts were checked to ensure an uncomplicated pregnancy. Uncomplicated pregnancies were defined
as pregnancies not complicated by gestational hypertension,
PE, HELLP syndrome or fetal growth restriction, placental
abruption or intrauterine fetal demise.
Patients and controls underwent the same cardiovascular
risk screening according to standardised protocol. At the
time of cardiovascular risk screening, all women were nonpregnant and had stopped breastfeeding, women who had
pregnancies after the index pregnancy had to be at least
6 months postpartum.
Exclusion criteria used in this study were known diabetes
mellitus, auto-immune diseases and pre-existent hypertension prior to index-pregnancy, as these diseases could lead
to bias. Finally, participants who did not wish to be
informed about the outcome of the screening were excluded.
Measurements
The cardiovascular risk screening started at 08:00 hours in
a temperature-controlled room (22°C), after an overnight
fast. Body weight (kg, Seca 888 scale, Hamburg, Germany)
and height (m) were measured. After 15 minutes’ rest,
blood pressure (BP) was determined for 30 minutes (at a
3-minute interval) in upright sitting position, using a semiautomatic oscillometric device (Dinamap Vital Signs Monitor 1846; Critikon, Tampa, FL, USA) with a cuff-size
appropriate for arm circumference. Participants were
instructed not to talk during measurements. The median
blood pressure was used for analysis. As the Framingham
risk model both weighs the use of antihypertensive
medication and actual blood pressure, women were
assigned to the hypertensive group when using anti-hypertensive medication. A venous blood sample was taken at
the level of the antecubital vein and analysed for fasting
glucose (mmol/l), insulin (mmol/l) and lipids (mg/dl):
low-density lipoprotein (LDL), high density lipoprotein
(HDL), triglycerides and total cholesterol. Body mass index
(BMI) was calculated by dividing body weight (kg) by
1093
Breetveld et al.
squared height (m). Participants filled out a questionnaire
regarding general history, current medication intake, intoxications, lifestyle factors and family history for CVD (in
first line relatives <60 years old). The cardiovascular risk
screening was performed following a standardised study
protocol by one experienced physician.
in an average 30-year Framingham Risk score (in healthy
women estimated to be 10–3.5%) between the PE and the
control group. A minimum of 42 patients per group was
needed for power of 80%, using an alpha of 5% for determining statistical significance. To compensate for possible
heterogeneity among formerly PE patients, we included at
least two formerly PE patients for each included control.
Framingham risk scores
Framingham risk scores were calculated with a gender-specific multivariable risk factor algorithm.20 Variables
included sex, age, systolic blood pressure (SBP), hypertension treatment, current smoking and de novo diabetes mellitus plus lipid spectrum (HDL and total cholesterol) or
body mass index (BMI).20–23 The 10-year risk for CVD was
calculated with the ‘Cardiovascular disease (10-year risk)’
calculator using BMI or lipids in the age range of 30–
74 years.21 Due to the young age of a few participants, the
risk score for 10 years could not be estimated for nine participants. The 30-year risk for CVD was calculated with the
‘Cardiovascular disease (30-year risk)’ calculator using BMI
or lipids in the age range of 20–59 years.22 Likewise, for
the 30-year risk, the ‘full CVD’ risk score was used which
included coronary death, myocardial infarction, fatal or
non-fatal stroke, coronary insufficiency, angina pectoris,
transient ischaemic attack, intermittent claudication or
congestive heart failure.22
Data analysis
We performed all statistical analyses using SPSS version
20.0 (version 20, IBM SPSS Statistics, Armonk, NY, USA).
We reported normally distributed continuous variables as
mean (standard deviation) and otherwise as median
(range). Binary variables were reported as absolute value
(percentage). Formerly preeclamptic patients were subdivided into two groups: a group with hypertension (HTPE)
at the time of the cardiovascular evaluation and women
without hypertension (NTPE). In addition, we subdivided
the PE group into a group with a history of early-onset of
PE and a group with a history of late PE. We used the
independent t-test for evaluating differences between
groups in continuous variables that were normally distributed. Dichotomous data were analysed using the chi-square
test. Differences between variables that were not normally
distributed were analysed using the nonparametric Kruskal–Wallis test or the nonparametric Mann–Whitney U-test
where appropriate. Differences between the Framingham
risk score based on BMI versus lipids were analysed with
the Wilcoxon Signed Rank test. A two-sided P-value of
<0.05 was considered statistically significant. We used the
difference in risk of developing CVD within 30 years after
gestation between the PE and control group to estimate the
sample size needed. We determined this sample size on the
basis of a clinically relevant proportional difference of 25%
1094
Results
In this observational study we included 115 patients with a
history of PE and 50 controls with an uncomplicated pregnancy. The characteristics of patients and controls are
shown in Table 1. Most women belonged to the European
continental ancestry group, two women were of Turkish
origin, and one was of Moroccan origin. Patients with previous PE, were on average, 3 years younger, had higher
weight and BMI and were using less alcohol compared with
controls. The interval between delivery at index pregnancy
and cardiovascular risk screening was approximately
2 years longer (P < 0.01) and gestational age at delivery
was approximately 6 weeks shorter in the PE group compared with controls (P < 0.01).
In the former PE group, 21 women had been diagnosed
with hypertension and 94 women had not been. In the control group, none of the women was diagnosed with chronic
hypertension. The variables of the Framingham risk calculator are presented in Table 2. Patients in the hypertensive
formerly PE group have a higher age, SBP and BMI
(P < 0.01). Moreover, early-onset PE clustered more often
in the hypertensive former PE group than in the normotensive former PE group, 90 and 61%, respectively (P < 0.01).
The results of the 10- and 30-year Framingham risk
scores are presented in Table 3. The 10-year predicted risk
score weighing subjects’ lipids did not differ between formerly PE patients and controls (1.6 and 1.5%, respectively;
P = 0.22), nor did the 30-year predicted risk score using
lipids (9.0 and 9.0%, respectively; P = 0.49). The 10-year
predicted risk score for the hypertensive formerly PE was
higher than in normotensive formerly PE women and controls (respectively, 3.1, 1.5 and 1.5%; P < 0.01). The 30year predicted risk was also higher in the hypertensive formerly PE than in the other groups (respectively, 19.0, 8.0
and 9.0%; P < 0.01). The 10-year predicted risk score using
BMI instead of lipid profile did not differ between formerly
PE patients and controls (1.5 and 1.5% respectively;
P = 0.60), nor did the 30-year predicted risk score (9.0 and
9.0% respectively; P = 0.86). In contrast, the 10-year predicted risk score for the hypertensive formerly PE was
higher than that in normotensive formerly PE and controls
(respectively, 3.4, 1.5 and 1.5%; P < 0.01). Also, the 30year predicted risk was higher in the hypertensive formerly
PE compared with the other groups (respectively, 19.0, 8.0
ª 2014 Royal College of Obstetricians and Gynaecologists
Framingham risk score in formerly preeclamptic women
Table 1. Characteristics of PE and CO group
Patient characteristics
Age, years
Height, m
Weight, kg
BMI, kg/m2
Obesity (BMI >30 kg/m2),
n (%)
Smoking, n (%)
Alcohol, n (%)
Family history of CVD,
n (%)
Obstetric variables
Primiparous, n (%)
Postpartum, years
Recurrent PE, n (%)
Index pregnancy
GA at birth, weeks
Birth weight, g
SGA, n (%)
IUFD, n (%)
Early onset PE, n (%)
Biochemical markers
Triglycerides, mmol/l
Total Cholesterol, mg/dl
LDL, mg/dl
HDL, mg/dl
Glucose, mmol/l
Insulin, mmol/l
HOMA, IR
Blood pressure
Systolic blood pressure,
mmHg
Diastolic blood pressure,
mmHg
Mean arterial pressure,
mmHg
Controls
(n = 50)
Formerly
PE (n = 115)
P-value
39 4.0
1.71 0.1
69 12
23.3 3.0
2/50 (4)
36 4.0
1.69 0.1
74 18
25.6 6.2
21/115 (18)
<0.01
0.09
<0.05
<0.01
<0.05
5/50 (10)
36/50 (72)
22/50 (44)
9/115 (8)
26/115 (23)
49/115 (43)
0.76
<0.01
0.87
5/50 (10)
8.0 2.7
–
41/115 (36)
5.4 2.6
27/82 (33)
<0.01
<0.01
–
39.6 2.3
3367 574
–
–
–
33.3 4.3
1800 939
61/115 (53)
8/115 (7)
76/115 (66)
<0.01
<0.01
–
–
–
0.9
187.0
110.6
61.6
4.7
6.3
1.3
0.4
29.7
25.9
13.0
0.4
3.6
0.8
1.0
180.1
112.7
51.8
4.8
8.8
1.9
0.5
28.6
24.5
9.9
0.6
4.7
1.1
0.15
0.16
0.62
<0.01
0.27
<0.01
<0.01
110 10
117 13
<0.01
71 7
74 10
0.07
82 8
86 11
<0.01
Significant values are written in bold.
BMI, body mass index; CVD, cardiovascular disease; GA, gestational
age; HDL, high density lipoprotein; HOMA, homeostatic model
assessment; IUFD, intrauterine fetal death; LDL, low density
lipoprotein; PE, preeclampsia; SGA, small for gestational age.
Table 4 present the results of the analysis after subdividing the PE group into a subgroup consisting of women
who had early-onset PE and a group of women who had
late-onset PE. The 10- and 30-year predicted risk score
weighing subjects’ lipids or BMI did not differ between
late-onset PE and controls. However, the 10- and 30-year
predicted risk score for early-onset PE based on lipids was
higher than in the late-onset PE group (respectively, 1.7
versus 1.3%; P < 0.05 and 10.0 versus 7.0%; P < 0.05). The
10-year predicted risk score based on the lipids was higher
in the early-onset PE than the controls (1.7 and 1.5%,
respectively; P < 0.05) but only differed from the CO.
When weighing BMI, the 10-year predicted risk score did
not differ between early-onset PE and both other groups.
However, the 30-year predicted risk score for early-onset
PE was higher than in the late-onset PE group (10.0 and
8.0%, respectively P < 0.05) but did not differ from the
control group.
When comparing both risk models (Table 4) we
observed a significantly higher score for the risk model
weighing BMI compared with the model weighing lipids in
the 10-year risk assessment for the late-onset PE group and
the CO group and for the 30-year risk assessment in the
CO group.
Next, we subdivided the former PE group into four subgroups based on onset and whether they had hypertension.
In the late-onset PE group, 37/39 (95%) were normotensive
and 2/39 (5%) were HT. In the early-onset PE group, 57/
76 (75%) were NT and 19/76 (25%) were hypertensive.
The 10- and 30-year predicted risk score based on the lipids was higher in the early PE-HT compared with the early
PE-NT and controls (3.0 versus 1.6 versus 1.5%, P < 0.01;
17 versus 9 versus 9%, P < 0.01, respectively). The 10- and
30-year predicted risk score based on the BMI was also
higher in earlyPE-HT compared to early PE-NT and controls (3.1% versus 1.5% versus 1.5% P < 0.01; 19% versus
8% versus 9% P < 0.01, respectively). However, the modelled cardiovascular disease risk of earlyPE-NT did not
differ in any way from that observed in controls.
Discussion
Main findings
and 9.0%; P < 0.01). The differences we observed were
irrespective of the model used, either with or without the
use of observed lipid profile (Table 3). Despite comparable
estimated median risk, comparing both risk models, the
10-year risk weighing BMI or lipid profile, we observed significant differences for the CO group and PE group. The
30-year risk models using BMI versus lipid profile indicated
significant differences for the CO group, but comparable
results for PE, NTPE and HTPE.
ª 2014 Royal College of Obstetricians and Gynaecologists
Despite the fact that formerly PE patients have elevated risk
of CVD and death compared to healthy parous controls,
still the largest fraction will not suffer from these remote
vascular complications. In order to differentiate between
those with low and elevated risk, we modelled remote cardiovascular risk using Framingham risk scores in formerly
PE patients and controls 5 to 10 years after birth, and
analysed the predicted 10- and 30-year risk scores. In contrast to our expectations, as group, formerly PE patients
had comparable risk estimates compared to healthy parous
1095
Breetveld et al.
Table 2. Demographic characteristics of CO, NTPE and HTPE groups
Age, years
Early onset PE, n (%)
Systolic Blood pressure, mmHg
Anti hypertensive treatment, n (%)
Smoking, n (%)
Diabetes, n (%)
BMI, kg/m2
HDL, mg/dl
Total cholesterol, mg/dl
CO (n = 50)
NTPE (n = 94)
HTPE (n = 21)
P-value
39 4.0
0 (0)
110 10
0 (0)
5/50 (10%)
0 (0)
23.3 3.0
61.6 13.0
187.0 29.7
35 3.9
57/94 (61%)
115 12
0 (0)
8/94 (9%)
0 (0)
25.1 6.2
52.2 10.3
178.0 28.8
37 3.8
19/21 (90%)
125 12
21/21 (100%)
1/21 (5%)
1/21 (5%)
28.0 5.9
50.0 7.8
189.9 26.5
<0.01
<0.01
<0.01
<0.01
0.85
0.13
<0.01
<0.01
0.11
Table 3. Predicted Framingham CVD risk scores based on lipids or BMI in formerly PE women and healthy parous controls
10-year CVD risk
Lipids
BMI
30-year CVD risk
Lipids
BMI
CO (n = 50)
PE (n = 115)
NTPE (n = 94)
HTPE (n = 21)
1.5 (1.0; 1.9)
1.5 (1.2; 2.3)***
1.6 (1.1; 2.4)
1.7 (1.2; 2.4)***
1.5 (1.1; 2.0)
1.5 (1.1; 2.0)
3.1 (2.2; 4.3)*,**
3.4 (2.2; 5.0)*,**
9.0 (7.0; 12.0)
9.0 (8.0; 14.0)***
9.0 (7.0; 14.0)
9.0 (7.0; 14.0)
8.0 (6.0; 11.0)
8.0 (6.8; 11.0)
19.0 (15.0; 26.0)*,**
19.0 (14.0; 28.5)*,**
*P < 0.01 relative to CO.
**P < 0.01 relative to NTPE.
***Significantly higher than the calculator using lipids.
Table 4. Predicted Framingham CVD risk scores based on early- and late-onset PE in formerly PE women and healthy parous controls
10-year CVD risk
Lipids
BMI
30-year CVD risk
Lipids
BMI
CO (n = 50)
Late-onset PE (n = 39)
Early-onset PE (n = 76)
1.5 (1.0; 1.9)
1.5 (1.2; 2.3)***
1.3 (0.9; 2.2)
1.5 (1.1; 2.2)***
1.7 (1.3; 2.5)*,**
1.7 (1.3; 2.6)
9.0 (7.0; 12.0)
9.0 (8.0; 14.0)***
7.0 (6.0; 12.0)
8.0 (6.0; 11.0)
10.0 (7.0; 14.8)**
10.0 (8.0; 14.8)**
*P < 0.05 relative to CO.
**P < 0.05 relative to late-onset PE.
***Significantly higher than the calculator using lipids.
controls. In contrast to hypertension, other risk factors that
are known to be increased after PE (e.g. cholesterol, BMI)
did not contribute substantially to the modelled increased
risk for CVD in our studied population. Nonetheless, treated hypertensive formerly PE patients had twofold risk on
CVD—compared to both normotensive formerly PE
patients and controls. The increased risk of CVD after PE
appears to be primarily related to blood pressure control
and not subclinical biochemical classical cardiovascular risk
1096
factors. This study is a translation of the increased risk
after PE to a predicting risk score to the patient.
Strengths and limitations
This study compares both formerly PE patients and healthy
parous controls at a time interval after birth sufficient to
be fully recovered. As such, both the effect of pregnancy
course and underlying risk factors could be weighted reliably. Until now, studies that analysed the Framingham risk
ª 2014 Royal College of Obstetricians and Gynaecologists
Framingham risk score in formerly preeclamptic women
score in a population did not make a distinction between
hypertensive and normotensive former preeeclamptic
women.16–18 In fact, normotensive former PE women and
women with a history of late-onset PE had comparable
modelled cardiovascular event risk with that observed in
controls. Women on antihypertensive medication showed a
significant increased risk score. This study has some limitations that need to be addressed. First, most women that
participated in this study belonged to the European continental ancestry group. Therefore our findings may not fully
be translational to other populations. Second, on average,
healthy parous controls were older than formerly PE. This
difference might have affected calculated risk scores, but
can be expected in higher rather than lower risk estimates.
Third, the control group was recruited by advertisement in
contrast to the former PE group which was a hospital
based recruitment. The women responding to advertisement on cardiovascular check, may somehow have a predisposition for CVD or may have subtle signs making them
question their CV risk. If the latest is true, the scores of the
control group may give an overestimation of the CV risk.
Fourth, it would be an interesting analysis to study in the
control group a normotensive and hypertensive group.
Unfortunately, our set of included controls does not give
us the opportunity to answer this. However, the prevalence
of chronic hypertension amongst young women with a history of normotensive pregnancies is extremely low and
underscores the suggestion that the increased risk of cardiovascular disease in formerly preeclamptic women primarily originates from high blood pressure. Fifth, the
Framingham risk score has not been validated in former
PE patients. It may be that the Framingham risk score is
not fully applicable for patients with previous PE. Even
though we were unable to substantiate an intrinsic detrimental effect of PE itself on remote cardiovascular health,
we cannot completely rule out PE to be an independent
risk factor for CVD that should be included in the risk prediction independent of other conventional risk factors.
The Framingham risk CV scoring system is originally
developed using older populations of men and women,20
and may not be directly applicable in our study population with an average age of 37 years. Thus the scorings
system may underestimate the risk score in our population. It would require a very long-term follow-up of a
large cohort of postpartum women to determine alternate
cut-offs for the 10-year and 30-year risk estimates for cardiovascular events. Further, by weighing traditional cardiovascular risk factors in patients without vascular
gestational problems, differences in risk scores primarily
seem to originate from chronic hypertension despite treatment, unless BP is really substantially lowered. As the
prevalence chronic hypertension rises up to approximately
fourfold within 15 years after giving birth,4 our data sug-
ª 2014 Royal College of Obstetricians and Gynaecologists
gest that the excess in CVD confines to those developing
chronic hypertension.
Interpretation
The Framingham risk score is the most compared risk calculator and widely used in North American countries.16
Other risk models for ischaemic heart disease (IHD) and
stroke are SCORE, CUORE and the Reynolds risk
score.16,24 In comparison, the Framingham risk score is the
only model which is able to model both 10- and 30-year
risk.23 Despite similarities in estimated outcome, results are
conflicting in precision and accuracy. Although some claim
the Framingham 10-year risk calculator to estimate close to
the actual observed risk,25 others state that the Framingham
risk model overestimates the risk of CVD,26,27 or underestimate it in young women. Previous studies claimed the calculated 10- and 30-years CV risk, based on the
Framingham risk calculator to be raised in former PE
patients compared to controls.16–18 However, these studies
did not report on the differences between BMI and lipid
risk model calculators. In our study, the BMI based model
seems to present higher risk estimates on the CVD risk in
the next 10–30 years.
Nowadays it is not known which postpartum interval
needs to be taken into account to rule out the pregnancyinduced alterations in the CV system. It is important to
wait a certain period to assure that most PE changes have
returned to a steady state. When it is likely that this state is
reached, the Framingham risk calculator may be applicable
to calculate patients individual risk score. The outcome of
the risk calculator may be considered a surrogate marker
for CV outcome. We expected, considering the reported
elevated chance on cardiovascular disease in these women,
a higher cardiovascular risk in formerly preeclamptic
women as a whole, especially when also taking lipid profile
into account. To our surprise, we did not see this. Only
after taking hypertension into account did a higher risk
estimate become visible. Therefore we think that, even
though we detailed group differences, our analysis underscores the use of the risk calculator in individual care. Follow-up of this population is necessary to validate
expectation with observation.
Hypertension strongly affects CVD risk.28 Moreover,
hypertension relates to significant chronic disability29 and
increases the risk of progression to chronic kidney disease
and obvious cardiovascular morbidity and mortality.30,31
Coronary heart disease is three times more frequent in
hypertensive than in normotensive individuals29 and even
an SBP ≥ 115 mmHg accounts for two-thirds of cerebrovascular diseases and almost half of ischaemic heart disease
cases.32 As such, there is no single factor except elevated
BP that plays a more important role in increasing
cardiovascular morbidity, mortality and overall mortality.29
1097
Breetveld et al.
Functionally, antihypertensive medication has the ability to
reverse or correct high blood pressure-induced structural
and functional alterations in large and small arteries.33
Clinically, the detrimental effects on cardiovascular health
can be reversed by antihypertensives.34 In young patients
(30–54 years), hypertension treatment resulted in a 41%
reduction of cerebrovascular events and 27% risk reduction
of fatal and non-fatal cardiovascular events.35 Meta-analysis
of 1 million participants demonstrated a linear association
between SBP and DBP and the risk of CVD mortality
(down to 115 mmHg and 75 mmHg).36 Every 10 mmHg
and 5 mmHg decrease of the usual (long-term average)
SBP and DBP, respectively, results in a 40% reduction of
risk of death from stroke and a 30% reduction in the risk
of IHD.36 Even a small reduction of 2 mmHg from the
usual SBP results in a 10% lower stroke mortality and
about a 7% lower mortality from IHD or other vascular
causes throughout middle age.36 Consequently, antihypertensives are one of the more cost-effective methods of
reducing premature cardiovascular morbidity and mortality.37 It should be noted that lifestyle interventions could
already be sufficient in patients with mildly elevated BP,
and should always be suggested when patients receive medication, as these may lower the necessary dosage.38 However, compliance with long-term healthy lifestyle
adjustments is extremely low.34 Based on average achieved
reduction in traditional cardiovascular risk factors imputed
in the several cardiovascular risk models, lifestyle adjustment may lead to a 4–13% reduction in CVD in patients
with a history of PE.24
We found in our study the highest prevalence of hypertension after early-onset PE, which is in line with previous
findings.39,40 The different subclassifications in our study
allowed us to conclude that chronic hypertension is a key
factor in predicting CVD risk later in life, even in the
highest risk subgroup of women with early-onset PE.
Moreover, those women with a history of early-onset PE
who were normotensive, had comparable risk estimates as
controls. Nevertheless, our results suggest that early-onset
PE is also associated with the highest risk for hypertension, in contrast to late-onset PE, and there should be
greater awareness of the necessity of cardiovascular followup.
In contrast to hypertension, other risk factors that are
known to be increased after PE (e.g. cholesterol, BMI) did
not contribute substantially to the modelled increased risk
for CVD in our studied population, either after early-onset
or after late-onset disease. Therefore, the observations in
this study, together with the available reported evidence,
indicate that the PE-related increased risk for CVD is probably increased by chronic hypertension. Nevertheless, it is
still unclear whether (1) PE in the absence of other risk
factors also predisposes independently to chronic HT and
1098
CVD and (2) PE magnifies the negative effects of other
subclinical risk factors on cardiovascular health, thus expediting the development of premature CVD.
Preventive strategies, early detection and corrective BP
treatment towards healthy reference values reduces the risk
of coronary heart failure (CHF)41 and may prevent or
delay the onset of costly CVD or renal disease.37 It is
important to manage a careful follow-up, as more than
half of known and treated hypertensive subjects still end
up with uncontrolled BP despite antihypertensive medicines.42 The follow-up of former PE patients is also incomplete, as more than one-third of these women do not have
their blood pressure followed-up after gestation or opportunistic blood pressure measurements taken when a former
PE patient visits the GP for other reasons.43 Apparently,
pregnancies complicated by a vascular disorder are still not
being fully recognised as a potential risk factor for developing chronic hypertension and CVD. Besides, there is as yet
no tailored risk score model available to assess a patient’s
individual risk for CVD after PE. Clinicians often perform
CV risk assessment on different ways. Most studies have
shown group differences for CV risk with an increased risk
in former PE women compared with controls. In our
study, we used the Framingham risk assessment to calculate the individual CV risk of patients. It is possible to
communicate the increased risk to the patient, and make
them more aware of the importance of blood pressure
measurements and control during follow-up. More importantly, the cardiovascular risk assessment does not stop
after one follow-up screening. A woman with no current
hypertension has an increased risk of becoming hypertensive in the following years. We stress the importance of
ongoing surveillance which at present we do not think is
practised consistently.
Conclusions
In this study, patients with a history of PE without
chronic hypertension have a comparable estimated future
risk to develop CVD events compared with patients with
uncomplicated pregnancies. In contrast, patients with previous PE who develop chronic hypertension in the first
decade after pregnancy have an approximately twofold
higher risk of developing CVD in the next 10–30 years.
One of five former PE patients who are currently hypertensive seem to be destined for a cardiovascular event.
Only former PE women with chronic hypertension and/or
early-onset PE have an increased risk score. Apparently,
hypertension is a very important and useful risk factor in
this specific population. Our findings stress again the
importance of CVD screening and follow-up in former PE
patients with a special focus on blood pressure measurement and treatment.
ª 2014 Royal College of Obstetricians and Gynaecologists
Framingham risk score in formerly preeclamptic women
Disclosure of interests
We have no conflict of interests to report.
Contribution to authorship
MJV, RRS, APD, WMH and MEAS led this study in Nijmegen
(Netherlands). MEAS and CGH suggested looking at the predicted Framingham risk score for developing cardiovascular
disease in the next 10–30 years in this specific population (PE
and CO). NMB performed the analysis and wrote the manuscript, under the supervision of CGH, SMJK and MEAS. The
manuscript was revised and approved by each author.
Details of ethics approval
The study protocol was approved by the Nijmegen Medical
Centre Medical Ethics Committee before patient enrolment
(NL32718.091.10). All subjects gave written informed consent before participation. The followed procedures were in
conformity with institutional guidelines and adhered to the
principles of the Declaration of Helsinki and Title 45, U.S.
Code of Federal Regulation, Part 46, Protection of Human
Subjects, Revised 13 November 2001, effective 13 December
2001.
Funding
None. &
References
1 Evans CS, Gooch L, Flotta D, Lykins D, Powers RW, Landsittel D,
et al. Cardiovascular system during the postpartum state in women
with a history of preeclampsia. Hypertension 2011;58:57–62.
2 Zhang J, Meikle S, Trumble A. Severe maternal morbidity associated
with hypertensive disorders in pregnancy in the United States.
Hypertens Pregnancy 2003;22:203–12.
3 ACOG practice bulletin. ACOG Committee on Practice Bulletins –
Obstetrics: Diagnosis and management of preeclampsia and
eclampsia. Obstet Gynecol 2002;99:159–67.
4 Bellamy L, Casas JP, Hingorani AD, Williams DJ. Pre eclampsia and
risk for cardiovascular disease and cancer in later life: systematic
review and meta analysis. BMJ 2007;335:974–7.
5 Melchiorre K, Sutherland GR, Liberati M, Thilaganathan G.
Preeclampsia is associated with persistent postpartum cardiovascular
impairment. Hypertension 2011;58:709–15.
6 Gifford RW, August PA, Cunningham G, Green LA, Lindheimer MD,
McNellis D, et al. National high blood pressure education program
working group on high blood pressure in pregnancy. Report of the
national high blood pressure education program working group on
high blood pressure in pregnancy. Am J Obstet Gynecol 2000;183:
S1–22.
7 Roger VL, Go AS, Lloyd-Jones DM, Benjamin EJ, Berry JD, Borden
WB, et al. Heart disease and stroke statistics – 2012 update: a
report from the American Heart Association. Circulation 2012;125:
e2–220.
8 Van Rijn BB, Nijdam ME, Bruinse HW, Roest M, Uiterwaal CS,
Grobbee DE, et al. Cardiovascular disease risk factors in women
with a history of early-onset preeclampsia. Obstet Gynecol
2013;121:1040–8.
ª 2014 Royal College of Obstetricians and Gynaecologists
9 Smith GN, Walker MC, Liu A, Wen SW, Swansburg M, Ramshaw H,
et al. A history of preeclampsia identifies women who have
underlying cardiovascular risk factors. Am J Obstet Gynecol
2008;200:58.e1–8.
10 Manten GTR, Sikkema MJ, Voorbij HAM, Visser GHA, Bruinse HW,
Franx A. Risk factors for cardiovascular disease in women with a
history of pregnancy complicated by preeclampsia or intrauterine
growth restriction. Hypertens Pregnancy 2007;26:39–50.
11 Hermes W, Franx A, van Pampus MG, Bloemenkamp KWM, Bots
ML, van dPJ, et al. Cardiovascular risk factors in women who had
hypertensive disorders late in pregnancy: a cohort study. Am J
Obstet Gynecol 2013;208:474.e1–e8.
12 Brown MC, Best KE, Pearce MS, Waugh J, Robson SC, Bell R.
Cardiovascular disease risk in women with pre eclampsia: systematic
review and meta analysis. Eur J Epidemiol 2013;28:1–19.
13 Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H,
Kannel WB. Prediction of coronary heart disease using risk factor
categories. Circulation 1998;97:1837–47.
14 National Health and Medical Research Council (NHMRC). Guidelines
for the assessment of absolute cardiovascular disease risk. 2009.
Available from: www.heartfoundation.org.au/SiteCollectionDocu
ments/guidelines-Absolute-risk.pdf. Accessed 2 January 2014.
15 Greenland P, Alpert J, Beller GA, Benjamin EF, Budoff MJ, Fayad ZA,
et al. 2010 ACCF/AHA Guideline for Assessment of Cardiovascular
Risk in Asymptomatic Adults: A Report of the American College of
Cardiology Foundation/American Heart Association Task Force on
Practice Guidelines. J Am Coll Cardiol 2010;56:e50–103.
16 Hermes W, Tamsma JT, Grootendorst DC, Franx A, van der Post J, van
Pampus MG, et al. Cardiovascular risk estimation in women with a
history of hypertensive pregnancy disorders at term: a longitudinal
follow-up study. BMC Pregnancy Childbirth 2013;13:126.
17 Fraser A, Nelson SM, Macdonald-Wallis C, Cherry L, Butler E, Sattar
N, et al. Associations of pregnancy complications with calculated
cardiovascular disease risk and cardiovascular risk factors in middle
age: the avon longitudinal study of parents and children. Circulation
2012;125:1367–80.
18 Smith GN, Pudwell J, Walker M, Wen S. Ten-year, thirty-year, and
lifetime cardiovascular disease risk estimates following a pregnancy
complicated by preeclampsia. J Obstet Gynaecol Can 2012;34:830–
5.
19 Brown MA, Lindheimer MD, de Swiet M, van Assche A, Moutquin
JM. The classification and diagnosis of the hypertensive disorders of
pregnancy: statement from the international society for the study of
hypertension in pregnancy (ISSHP). Hypertens Pregnancy 2001;20:9–
14.
20 D’Agostino RB Sr, Vasan RS, Pencina MJ, Wolf PA, Cobain M,
Massaro JM, et al. General cardiovascular risk profile for use in
primary care: the Framingham Heart Study. Circulation
2008;117:743–53.
21 Framingham Heart Study. General Cardiovascular Disease (10-year
risk) (based on D’Agostino RB, Vasan RS, Pencina MJ, Wolf PA,
Cobain M, Massaro JM, et al. General cardiovascular risk profile for
use in primary care: The Framingham Heart Study. Circulation
2008;117:743–53.) 2014 Available from: www.framingham
heartstudy.org/risk-functions/cardiovascular-disease/10-year-risk.php.
Accessed 19 May 2014.
22 Framingham Heart Study. Cardiovascular Disease (30-year risk)
(based on Pencina MJ, D’Agostino RB, Larson MG, Massaro JM,
Vasan RS. Predicting the 30-year risk of cardiovascular disease.
The Framingham Heart Study. Circulation 2009;119:3078–3084.)
2014
Available
from:
www.framinghamheartstudy.org/riskfunctions/cardiovascular-disease/30-year-risk.php. Accessed 19 May
2014.
1099
Breetveld et al.
23 Pencina MJ, D’Agostino RB Sr, Larson MG, Massaro JM, Vasan RS.
Predicting the 30-year risk of cardiovascular disease: The
Framingham Heart Study. Circulation 2009;119:3078–84.
24 Berks D, Hoedjes M, Raat H, Duvekot JJ, Steegers EAP, Habbema
JDF. Risk of cardiovascular disease after pre-eclampsia and the effect
of lifestyle interventions: a literature – based study. BJOG
2013;120:924–31.
25 Sayin MR, Cetiner MA, Karabag T, Akpinar I, Sayin E, Kurcer MA,
et al. Framingham risk score and severity of coronary artery disease.
Herz 2013;1–6.
26 Barroso LC, Muro EC, Herrera ND, Ochoa GF, Hueros JIC, Buitrago
F. Performance of the framingham and SCORE cardiovascular risk
prediction functions in a non-diabetic population of a Spanish health
care centre: a validation study. Scand J Prim Health Care
2010;28:242–8.
27 Giavarina D, Barzon E, Cigolini M, Mezzena G, Soffiati G.
Comparison of methods to identify individuals at increased risk of
cardiovascular disease in Italian cohorts. Nutr Metab Cardiovasc Dis
2007;17:311–18.
n J, Zanchetti A, Bo
€hm M,
28 Mancia G, Fagard R, Narkiewicz K, Redo
et al. 2013 Practice guidelines for the management of arterial
hypertension of the European Society of Hypertension (ESH) and the
European Society of Cardiology. J Hypertens 2013;31:1925–38.
29 Mancia G, Grassi G. Management of essential hypertension. Br Med
Bull 2010;94:189–99.
30 Lim S. Recent update in the management of hypertension. Acta
Med Indones 2007;39:186–91.
31 Israili ZH, Hern
andez-Hernandez R, Valasco M. The future of
antihypertensive treatment. Am J Ther 2007;14:121–34.
32 World Health Organization. The World Health Report 2002:
Reducing Risks, Promoting Healthy Life. Geneva: WHO, 2002.
33 Rehman A, Schiffrin EL. Vascular effects of antihypertensive drug
therapy. Curr Hypertens Rep 2010;12:226–32.
1100
34 Mancia G, De Backer G, Dominiczak A, Cifkova R, Fagard R,
Germano G, et al. 2007 Guidelines for the management of arterial
hypertension. J Hypertens 2007;25:1105–87.
35 Quan A, Kerlikowske K, Gueyffier F, Boissel JP. Efficacy of treating
hypertension in women. J Gen Intern Med 1999;14:718–29.
36 Lewington S, Clarke R, Qizilbash N, Peto R, Collins R; Prospective
Studies Collaboration. Age-specific relevance of usual blood
pressure to vascular mortality: a meta-analysis of individual data for
one million adults in 61 prospective studies. Lancet 2002;360:
1903–13.
37 Elliott WJ. The economic impact of hypertension. J Clin Hypertens
(Greenwich) 2003;5(3 suppl 2):3–13.
38 Perk J, de Backer G, Gohlke H, Graham I, Reiner Z, Verschuren
WMM, et al. European Guidelines on cardiovascular disease
prevention in clinical practice (version 2012). Eur Heart J 2012;33:
635–1701.
39 Melchiorre K, Thilaganathan B. Maternal cardiac function in
preeclampsia. Curr Opin Obstet Gynecol 2010;23:440–7.
40 Ghossein-Doha C, Peeters L, van Heijster S, van Kuijk S, Spaan J,
Delhaas T, et al. Hypertension after preeclampsia is preceded by
changes in cardiac structure and function. Hypertension
2013;62:382–90.
41 Levy D, Larson MG, Vasan RS, Kannel WB, Ho KK. The progression
from hypertension to congestive heart failure. JAMA 1996;275:
1557–62.
42 Schelleman H, Klungel OH, Kromhout D, de Boer A, Stricker
B. Prevalence and determinants of undertreatment of
hypertension in the Netherlands. J Hum Hypertens
2004;18:317–24.
43 Nijdam ME, Timmerman MR, Franx A, Bruinse HW, Numans
ME, Grobbee DE, et al. Cardiovascular risk factor assessment
after pre-eclampsia in primary care. BMJ Fam Pract 2009;
10:77.
ª 2014 Royal College of Obstetricians and Gynaecologists